Hi - see below.
> On 26 Oct 2016, at 15:58, Rodrigo Perea <[log in to unmask]> wrote:
>
> Hi FSL community,
>
> <<
> $cat /usr/pubsw/packages/fsl/current/etc/fslversion
> 5.0.7
>>>
>
> ##So i run the following probtrackx command:
>
> ${FSL_DIR}/current/bin/probtrackx2 --network -x ${OUTPUTFOLDER}/masks.txt -l --onewaycondition -c 0.2 -S 2000 --steplength=0.5 -P 5000 --fibthresh=0.01 --distthresh=0.0 --sampvox=0.0 --forcedir --opd -s ${BEDPOSTX_FOLDER}/merged -m ${BEDPOSTX_FOLDER}/nodif_brain_mask --dir=${OUTPUTFOLDER} --otargetpaths
>
> *My ${OUTPUTFOLDER}/masks.txt contains the location of 8 masks.
>
>
> ##My output files in ${OUTPUTFOLDER}/* looks like this:
>
> fdt_network_matrix fdt_paths.nii.gz masks.txt probtrackx.log tmpnetmaskfile waytotal
>
> *The waytotal file ($cat waytotal):
> 494180
> 818570
> 701471
> 658294
> 1150165
> 918453
> 404775
>
> *The fdt_network_matrix ($cat fdt_network_matrix):
> 0 71666 41647 748 75844 58106 805
> 181289 0 101714 100823 22199 13932 1749
> 14919 158258 0 1140 82435 111302 705
> 12783 298912 7626 0 4805 3476 19875
> 236934 42600 56841 2551 0 193554 69016
> 49653 35167 149099 482 228130 0 37352
> 6920 8388 1490 27710 145783 27243 0
>
>
>
> ##My questions:
>
> 1) I assume fdt_network_matrix is a ROI-by-ROI matrix that specifies the streamlines that get generated from each ROI to another and is not symmetric because the starting seeding ROI will vary the number of streamlines that reach each other ROI, correct?
> *I couldn't find any explicit information on what this output is [ not even here: http://fsl.fmrib.ox.ac.uk/fsl/fslwiki/FDT/UserGuide#Overview ]...could you please confirm what I believe is happening?)
Correct. It is also asymmetric because the tracking algorithm itself is not symmetric (imagine a Y shaped tract, coming from the branches keeps 100% of tracts, coming from the trunk splits the connections).
>
> 2) Based on what I know waytotal is the number of generated tracts that aren't rejected. If my number of samples is "-P 5000", why am I getting such larger values in the waytotal output?
Because 5000 is the number of samples per seed voxel. You need to multiply that by the number of seed voxels to get the total number of samples.
>
>
> 3) What is the relationship between fdt_network_matrix and waytotal?
Given the answer to (2) above, I think this should be clearer now.
>
> 4) I added the hidden option --otargetpaths but no additional paths were generated. I've used this option before but no paths were generated. Is this hidden option still working?.
This option is meant to go with the —targetmasks option, i.e. it produces one fdt_paths per target mask. We do not have something equivalent for the network option, but you should be able to produce those by running probtrackx2 separately for each seed ROI.
>
> Thanks in advance,
>
> Rodrigo
|